CLAIMS
What is claimed is:
1. A method (500) of selective encryption of a test dataset, the method comprising:
determining (504), by an encrypting device (102), a relevancy grade associated with each of a plurality of datapoints within a test dataset by comparing the test dataset with a common heat map, wherein the common heat map is generated using a plurality of training datasets;
calculating (506), by the encrypting device (102), based on the relevancy grade, an encryption level associated with each of the plurality of datapoints; and
selectively encrypting (508), by the encrypting device (102), at least one datapoint from the plurality of datapoints based on the encryption level associated with each of the plurality of datapoints, wherein the at least one data point is rendered to a user after being decrypted.
2. The method of claim 1, wherein generating (600) the common heat map comprises:
receiving (602) the plurality of training datasets, each associated with a classification from one or more classifications, wherein the one or more classifications are determined by an Artificial Intelligence (Al) model;
generating (604) one or more heat maps for one or more of the plurality of training datasets, wherein each of the one or more heat maps indicates a relevancy grade associated with each datapoint of each of the one or more training datasets; and
superimposing (610) the one or more heat maps for the one or more training datasets, to generate the common heat map.

3. The method of claim 1, wherein determining (504) the relevancy grade further
comprises:
receiving (606) one or more filters used by the Al model for determining the one or more classifications for the plurality of training datasets; and
determining (608) a relevancy grade associated with each of the plurality of datapoints of the test dataset based on the one or more filter.
4. The method of claim 1 further comprises:
assigning (502) a priority index to each of a plurality of regions of the common heat map, based on relative importance of the respective region; and
determining (504) the relevancy grade associated with each of a plurality of datapoints within a test dataset based on the priority index assigned to each of the plurality of regions of the common heat map, wherein selectively encrypting (508) the at least one datapoint is based on an iterative encrypting (516) each of the at least one datapoint, and wherein a number of iterations of the encryption are based on the relevancy grade associated with each of the at least one datapoint.
5. The method of claim 4, wherein selectively encrypting (508) the at least one datapoint
comprises:
selecting (518) a homomorphic encryption key of a predetermined complexity for encrypting each of the at least one datapoint based on the relevancy grade assigned to each of the at least one datapoint; and

encrypting (520) each of the at least one datapoint using the associated selected homomorphic encryption key, and
generating (510) an encryption scale based on encrypted datapoints of the test dataset and unencrypted datapoints of the test dataset.
6. The method of claim 5 further comprising:
decrypting (512) an encrypted classification output based on the encryption scale to generate a decrypted classification output; and
rendering (514) the decrypted classification output to the user in response to the decryption, wherein the decrypted classification output comprises the at least one datapoint.
7. An encryption device (102) for selective encryption of a test dataset, the encryption
device (102) comprising:
a processor (110); and
a memory (112) communicatively coupled to the processor (110), wherein the memory (112) stores processor instructions, which, on execution, causes the processor (110) to:
determine (504) a relevancy grade associated with each of a plurality of
datapoints within a test dataset by comparing the test dataset with a common
heat map, wherein the common heat map is generated using a plurality of
training datasets;

calculate (506) based on the relevancy grade, an encryption level associated with each of the plurality of datapoints; and
selectively encrypt (508) at least one datapoint from the plurality of datapoints based on the encryption level associated with each of the plurality of datapoints, wherein the at least one data point is rendered to a user after being decrypted.
8. The encryption device (102) of claim 7, wherein generating (600) the common heat
map comprises:
receiving (602) the plurality of training datasets, each associated with a classification from one or more classifications, wherein the one or more classifications are determined by an Artificial Intelligence (Al) model;
generating (604) one or more heat maps for one or more of the plurality of training datasets, wherein each of the one or more heat maps indicates a relevancy grade associated with each datapoint of each of the one or more training datasets; and
superimposing (610) the one or more heat maps for the one or more training datasets, to generate the common heat map.
9. The encryption device (102) of claim 7, wherein determining (504) the relevancy
grade further comprises:
receiving (606) one or more filters used by the Al model for determining the one or more classifications for the plurality of training datasets; and

determining (608) a relevancy grade associated with each of the plurality of datapoints of the test dataset based on the one or more filter.
10. The encryption device (102) of claim 7, wherein the processor instructions further cause the processor to:
assign (502) a priority index to each of a plurality of regions of the common heat map, based on relative importance of the respective region;
determine (504) the relevancy grade associated with each of a plurality of datapoints within a test dataset based on the priority index assigned to each of the plurality of regions of the common heat map;
iteratively encrypt (516) each of the at least one datapoint, wherein a number of iterations is based on the relevancy grade associated with each of the at least one datapoint;
select (518) a homomorphic encryption key of a predetermined complexity for encrypting each of the at least one datapoint based on the relevancy grade assigned to each of the at least one datapoint;
encrypt (520) each of the at least one datapoint using the associated selected homomorphic encryption key, and
generate (510) an encryption scale based on encrypted datapoints of the test dataset and unencrypted datapoints of the test dataset.